School of Biological Sciences, University of Southampton,
Southampton SO16 7PX, United Kingdom
 |
INTRODUCTION |
The coordinated synthesis of
chlorophylls and chlorophyll-binding proteins is critically important
during emergence of the etiolated seedling into light, and is
essential for the normal development of functional chloroplasts. The
key regulatory step in chlorophyll synthesis is the formation of
5-aminolevulinic acid (ALA), which is rate limiting for the
tetrapyrrole pathway (Beale and Castelfranco, 1974
). As a
consequence, inhibition of ALA synthesis through inhibitors or
anti-sense approaches (Höfgen et al., 1994; Kumar and Söll,
2000) has a dramatic effect on chloroplast development, resulting in
plants that are highly susceptible to photobleaching. In a converse
manner, excess ALA also affects chloroplast development (Kittsteiner et
al., 1991) and leads to an overaccumulation of porphyrins, which can
lead to severe photooxidative damage (Reinbothe et al., 1996).
Uncoupling tetrapyrrole synthesis from chloroplast development may also
be achieved by irradiating etiolated seedlings with far-red (FR) light
(Barnes et al., 1996; Runge et al., 1996). Growth of Arabidopsis
seedlings under continuous FR (FRc) can induce partial
photomorphogenesis with reduced hypocotyl elongation growth and
cotyledon expansion. These responses are characteristic of the
FR-high-irradiance response (FR-HIR; Mancinelli, 1994), which is a
specific function of phytochrome A (Smith, 1995). FRc irradiation can
also activate many of the processes required for chloroplast
development. These include the induction of nuclear genes encoding
chlorophyll a/b-binding proteins and other
photosynthetic proteins (Kuno et al., 2000; Tepperman et al., 2001) and
the transcription of chloroplast genes and replication of plastid DNA
(DuBell and Mullet, 1995a, 1995b; Chun et al., 2001). However,
because photoconversion of protochlorophyllide (Pchlide) to
chlorophyllide (Chlide) by the enzyme NADPH:Pchlide
oxidoreductase (POR) is inefficient under FR wavelengths,
chlorophyll synthesis is severely limited under these conditions. Thus,
phytochrome A-mediated induction of selected facets of plastid
development may occur in the absence of conditions that allow the
synthesis of corresponding levels of chlorophyll.
It has previously been shown that seedlings of Arabidopsis and tomato
(Lycopersicon esculentum) grown under prolonged FR
not only fail to accumulate chlorophyll, but also are unable to green when subsequently transferred to white light (WL; Barnes et al., 1995;
van Tuinen et al., 1995). This phenomenon, known as the FR block of
greening response, has been characterized at the molecular level as a
depletion of PORA (and partially of PORB) and the concomitant loss of
the ordered membrane system of the prolamellar body (Barnes et al.,
1996; Runge et al., 1996). Each of these effects displays characteristics of an FR-HIR and is dependent on phytochrome A activity
(Barnes et al., 1996). Further evidence that the loss of POR is crucial
to the FR block of greening comes from a transgenic approach in which
overexpression of PORA was able to maintain WL-mediated
greening in FR-pretreated seedlings (Runge et al., 1996; Sperling et
al., 1997). It is likely that PORA has two important roles in
maintaining chlorophyll synthesis in WL. Not only is it required for
light-dependent chlorophyll synthesis, but it has a role in buffering
against photooxidative damage that occurs during the illumination of
etiolated, Pchlide-containing tissues (Runge et al., 1996). This second
role has been largely supported by subsequent studies with
PORA- and PORB-overexpressing lines (Sperling et
al., 1997, 1999).
However, it is interesting that POR overexpression could not
fully counteract the FR block, and FR-pretreated transgenic seedlings still contained approximately two-thirds of the chlorophyll levels of
dark-treated controls (Sperling et al., 1997). This suggests that
although the level of POR is a critical determinant of this response,
other FR-mediated processes may also be important. Prolonged FR
irradiation has been shown to cause the formation of aberrant plastids
in developing cotyledons, and the irreversible nature of the FR block
of greening response has been attributed to the ensuing degradation of
plastid components suggested by the formation of vesicles in the stroma
(Barnes et al., 1996). One possibility is that this FR-mediated effect
on plastid integrity leads to the loss of a plastid signal required for
normal expression of photosynthetically related genes. Such a signal
has been previously identified through the action of photobleaching
herbicides (Taylor, 1989; Susek and Chory, 1992) and is required for
the expression of numerous photosynthetically related nuclear genes
(Kusnetsov et al., 1996), including Lhcb in Arabidopsis
(Susek et al., 1993). This signal is also closely involved with
phytochrome signaling pathways (López-Juez et al., 1998) and can
be thought of as gating phytochrome responses (McCormac et al., 2001)
in a similar way to that proposed for circadian control of
light-induced gene expression (Millar and Kay, 1996).
We have examined the hypothesis that plastid signaling is involved in
the FR block of greening response by analyzing the expression profile
of HEMA1. This gene, which encodes glutamyl-tRNA reductase, the first committed enzyme of tetrapyrrole synthesis, is strongly expressed in photosynthetic tissues and is regulated by light, including an FR-HIR, and a plastid signal (Ilag et al., 1994; Kumar et
al., 1999; McCormac et al., 2001). However, the light-regulated expression pattern of HEMA1 is not identical to other
light-regulated genes, as HEMA1 does not exhibit a very
low-fluence response (McCormac et al., 2001) and utilizes light
signaling pathways that partly diverge from those for light regulation
of Lhcb (A.C. McCormac and M.J. Terry, unpublished
data). Given the importance of tetrapyrroles in the FR block of
greening response and the increasingly significant relationship between
tetrapyrroles and plastid signaling (Kropat et al., 2000; Vinti et al.,
2000; La Rocca et al., 2001; Mochizuki et al., 2001; Møller et
al., 2001), the expression of HEMA1 under these conditions
is likely to be highly informative for our understanding of plastid development.
| |
RESULTS |
The FR Block of Greening Response Is Associated with an
Inhibition of Promoter Responsiveness to WL
Figure 1A shows that, as seen
previously (Barnes et al., 1996; Runge et al., 1996), 3 d of FRc
(10 µmol m
2 s1)
irradiation of etiolated seedlings fully inhibited their ability to
green under subsequent WL, whereas dark-grown seedlings at the same
developmental age could green normally. This effect was independent of
WL intensity (Fig. 1A; Barnes et al., 1996) and was also found to be
associated with a loss of HEMA1 and Lhcb expression (3 d FRc; Fig. 1B). Seedlings that had received a
pretreatment of just 1 d of FRc suffered only a small, but
consistently observed, loss of greening capacity that was enhanced by
high-intensity WL (Fig. 1A). These seedlings were still able to
accumulate high levels of HEMA1 and Lhcb mRNA (1 d FRc; Fig. 1B). Under these conditions, FRc stimulated subsequent WL
induction of HEMA1 expression by up to 2-fold compared with
seedlings transferred directly from darkness to WL (Fig. 1B). This
stimulation was most evident under low-intensity (8 µmol
m2 s1) WL, but was
absent under high-intensity (250 µmol m2
s1) WL (data not shown). These results were
confirmed using transgenic Arabidopsis seedlings expressing a
HEMA1 promoter:gusA construct (Fig. 1C).
Furthermore, the 1-d FRc-mediated stimulation of WL-induced HEMA1 expression and the 3-d FRc-mediated inhibition of
HEMA1 expression following transfer to WL were absent in a
phyA genetic background, indicating that these are both
phytochrome A-mediated responses (Fig. 1C). Therefore, the FR block of
greening response is characterized not only as a progressive loss of
greening ability, but also by a marked change in the transcriptional
response to WL.
View larger version (27K):
[in this window]
[in a new window]
|
Figure 1.
FRc preirradiation leads to a reduction in
greening and nuclear gene expression following transfer to WL. A,
Greening of cotyledons under different WL fluence rates after growth in
the dark or FRc. Seedlings were germinated without Suc in the dark for
2 d. They were subsequently grown for a further period (as
indicated) under FRc (white symbols) or were maintained in the dark
(black symbols) for the equivalent period. Seedlings were then
transferred to a WL source of 250 µmol m2
s1 (circles) or 8 µmol
m2 s1 (triangles). B,
HEMA1 and Lhcb mRNA accumulation under WL (130 µmol m2 s1) following
a 1- or 3-d FRc pretreatment. Seedlings were grown without Suc. The RNA
gel blot was sequentially hybridized with probes for HEMA1,
Lhcb, and 18S transcripts. C, Effect of an FR
pretreatment on the subsequent response to WL of a HEMA1
promoter: -glucuronidase A (gusA) transgene.
Transgenic WT or phyA Arabidopsis seedlings were germinated
for 1 d in the dark followed by a 1- or 3-d FRc
treatment (or equivalent darkness). Seedlings were then transferred to
WL (8 µmol m2 s1) for
3 d prior to measurement of GUS activity. Data shown are the
mean ± SE (n = 3).
|
|
The FR Block of Greening Response Develops without Inhibition of
HEMA1 Transcription in the Presence of Suc
A full FR block of greening response was avoided if seedlings were
grown on a medium supplemented with 3% (w/v) Suc (Fig. 2A; Barnes et al., 1996). However, even
in the presence of Suc, FR pretreatments still elicited a partial
impairment of the greening response as compared with that in dark-grown
seedlings (Fig. 2A). This partial block of greening was not seen
following FR irradiation of phyA lines (data not shown). In
contrast to the FR block of greening response in the absence of Suc
(Fig. 1A), the partial FR block in wild-type (WT) lines was shown to be
dependent on WL intensity (Fig. 2A). In addition, for seedlings grown
on 3% (w/v) Suc, an FR pretreatment produced no discernible inhibition of HEMA1 or Lhcb transcript accumulation
following transfer to 130 µmol
m2 s1 WL (Fig. 2B),
even with a 75% inhibition of greening capacity. The accumulation of
GUS activity driven by the HEMA1 promoter also showed a
normal response to WL following a 3-d FRc preirradiation and was even
elevated compared with the response of dark-grown seedlings in
low-intensity WL (data not shown). Thus, under these conditions, the
effect of an FR preirradiation was to separate HEMA1
expression from chlorophyll accumulation.
View larger version (41K):
[in this window]
[in a new window]
|
Figure 2.
Suc inhibits the loss of nuclear gene expression
in FRc-grown seedlings transferred to WL. A, Greening of cotyledons
under different WL fluence rates after growth in darkness or FRc.
Seedlings were germinated on 3% (w/v) Suc in the dark for 2 d.
They were subsequently grown for a further period (as indicated) under
FRc (white symbols) or were maintained in darkness (black symbols) for
the equivalent period. Seedlings were then transferred to a WL source
of 250 µmol m2 s1
(circles) or 8 µmol m2
s1 (triangles). B, HEMA1 and
Lhcb mRNA accumulation under WL (130 µmol
m2 s1) following a 1- or 3-d FRc pretreatment. Seedlings were grown on 3% (w/v) Suc. The RNA
gel blot was sequentially hybridized with probes for HEMA1,
Lhcb, and 18S transcripts.
|
|
The response characteristics on Suc are similar to those seen in the
absence of Suc, but following a short period (1 d) of FR irradiation
(Fig. 1, A and B) and they enable us to define two separate responses
leading to an FR block of greening. The first response requires only
1 d of FRc and results in a WL intensity-dependent, incomplete
loss of greening ability that is not associated with a loss of
HEMA1 and Lhcb expression. We subsequently refer
to this as the transcriptionally uncoupled response. The second
response requires a longer period of FRc (3 d), is independent of WL
intensity, and leads to a complete loss of greening ability (<10%)
and the inhibition of HEMA1 and Lhcb expression.
This WL intensity-independent response is specifically inhibited by Suc
and is now referred to as the transcriptionally coupled response.
Different Transcriptional Responses under FRc Define
the Two FR Block of Greening Responses under WL
To further define the characteristics of the two response pathways
leading to the FR block of greening, we examined the expression of
HEMA1 during FRc. Figure 3A
shows a time course of HEMA1 and Lhcb expression
for a 4-d period of FR irradiation. A peak in HEMA1 and
Lhcb mRNA levels occurred at around 1 d from the start of the FR treatment and declined thereafter, returning almost to dark
levels by d 4 (Fig. 3, A and B). This type of short-lived expression
profile is comparable with that reported previously for a wide range of
plastid-associated nuclear genes under FRc (Tepperman et al., 2001),
but is in contrast to the sustained transcriptional response seen under
Rc (Fig. 3C). The post-peak phase of FR-mediated HEMA1
expression (i.e. 
72 h FRc; Fig. 3A) temporally coincides with the
inability to reinitiate transcription when exposed to WL (3 d FRc; Fig.
1B). Because Suc can prevent the loss of HEMA1 expression in
WL after a prolonged FRc treatment (Fig. 2B), we tested whether it
could also block the loss of FR-induced HEMA1 expression.
Figure 3D shows that the loss of HEMA1 expression under FRc
was prevented by the presence of 3% (w/v) Suc with maximal expression
levels (detected after 3 d of FRc) remaining for 7d. However,
the presence of 3% (w/v) Suc also substantially reduced the
accumulation of HEMA1 (and Lhcb; data not shown)
mRNA seen within the first 24 h of FRc (Fig. 3, compare A and D).
This is consistent with previous reports of the effect of Suc on
phytochrome A signaling (Dijkwel et al., 1997).
View larger version (24K):
[in this window]
[in a new window]
|
Figure 3.
The response of the HEMA1 promoter to
light is subject to developmental aging, FR fluence rate, and Suc. A,
Transcript levels of HEMA1 and Lhcb under
prolonged (16-96 h) FR irradiation. Seedlings were grown without Suc
and were germinated for 2 d in the dark before transfer (at
time = 0) to FRc. B, Densitometry scans of band intensities as
shown in (A). C, HEMA1 transcript levels under prolonged
(24-96 h) R irradiation. Seedlings were grown without Suc and were
germinated for 2 d in the dark before transfer (at time = 0) to Rc. D, HEMA1 mRNA levels
in seedlings grown under prolonged (1-9 d) FR irradiation in the
presence of 3% (w/v) Suc. Seedlings were germinated for 3 d in
the dark prior to transfer (at t = 0) to FRc. E, HEMA1
mRNA levels in seedlings grown under 1 or 10 µmol
m2 s1 FRc. F,
HEMA1 mRNA levels under WL in seedlings grown previously for
3 d in the dark or under 1 or 10 µmol m2
s1 FRc. G, Effect of aging on the subsequent
response of a HEMA1 promoter:gusA construct to
WL. Seedlings were grown with or without 3% (w/v) Suc in the dark and
were subsequently transferred to WL for 3 d prior to measurement
of GUS activity. Data shown are the mean ± SE (n = 5).
|
|
Phytochrome A-mediated FR responses are also typically affected by the
rate of FR delivery (Mancinelli, 1994). Therefore, we examined this
relationship for the transcriptional response to FR and the FR block of
greening response by comparing the effect of irradiating seedlings
(grown without Suc) with FRc at 1 µmol m2
s1 (Fig. 3E) instead of 10 µmol
m2 s1 (Fig. 3, A and
E). Under the lower fluence rate, the levels of HEMA1 mRNA
showed the same pattern of transient up-regulation, but the rate of
transcriptional increase and magnitude of the peak were reduced (Fig.
3E). A 3-d pretreatment under 1 µmol m2
s1 FRc failed to induce the FR-mediated block
of the WL transcriptional response (Fig. 3F). Instead, the
FR-irradiated seedlings displayed the characteristics of the
transcriptionally uncoupled FR block of greening, showing strong
nuclear transcription (Fig. 3F) and a WL intensity-dependent loss of
greening (data not shown). In each case, phyA mutants
demonstrated that there was an absolute requirement for phytochrome A
for the response to FR (data not shown).
To test whether the decline in mRNA accumulation under FRc (Fig. 3B;
24-96 h) could be explained as a general loss of light responsiveness
rather than the specific loss of phytochrome A-mediated signaling, we
examined the effect of prolonging the period in darkness on the ability
of seedlings to induce HEMA1 in response to WL. There was a
progressive loss of HEMA1 light responsiveness to WL (Fig.
3G) and also FR (data not shown). This effect was also seen in
phyA mutants, which demonstrates that this is a
time-dependent, but phytochrome A-independent, response (data not
shown). In addition, a time-dependent depletion of greening in
dark-grown seedlings was seen in the absence of Suc (Fig. 1A). When 3%
(w/v) Suc was included in the growth medium, the time-dependent loss of
promoter responsiveness to WL (Fig. 3G) and FR (data not shown) was
prevented, suggesting that starvation may be a key factor in this
process. Thus, it appears that in darkness or FR light (in the absence of exogenous sugars), nuclear genes progress into a nonresponsive state. By contrast, under R (Fig. 4C) or
WL, the transcriptional response escapes this process.
View larger version (23K):
[in this window]
[in a new window]
|
Figure 4.
Relationship between HEMA1 expression,
ALA synthesis, and greening ability in FRc-grown seedlings. A, ALA
synthesis rates in seedlings of WT and the phyA mutant grown
in the dark (black columns) or under 1 d of FRc (hatched columns),
with or without 3% (w/v) Suc (as indicated). The inset graph shows the
corresponding levels of Pchlide accumulated in the FR-irradiated WT and
phyA seedlings. Data shown are the mean ± SE (n = 3). B, Relationship
between relative HEMA1 mRNA levels (as measured by
densitometry scans of RNA gel blots) and corresponding rates of ALA
synthesis in WT and phyA (inset) seedlings. Seedlings were
allowed to germinate in the presence of 3% (w/v) Suc for 2 d in
the dark and were transferred to FRc or maintained in the dark for an
additional period of 1 or 3 d. Each datum point shows the
individual value under either irradiation condition for these two
incubation times for the Landsberg erecta (Ler)
and Colombia (Col) backgrounds. C, Relationship between greening
capacity measured after transfer to WL and ALA synthesis rates prior to
transfer in dark- and FRc-grown WT ( ) and phyA ( )
seedlings. Seedlings were grown as for B. Data shown are the mean ± SE (n = 3).
|
|
These results demonstrate that a transcriptionally coupled block of
greening response can proceed in the absence of light, although more
slowly than under high-intensity FRc (see also Fig. 1A) and therefore
appears to have no absolute requirement for phytochrome A. In contrast,
the transcriptionally uncoupled effect on greening has a strict
requirement for phytochrome A at the seedling stage (see also Fig. 2A).
In addition, the two FR block of greening responses, as observed in WL,
can be further distinguished by the different patterns of
transcriptional response during the preceding period of FR
irradiation. Thus, for FR irradiation to accelerate the
development of the transcriptionally coupled block of greening
response, a maximal FR-HIR must be achieved. By contrast, a submaximal
FR-HIR can trigger the transcriptionally uncoupled block of greening.
The Transcriptionally Uncoupled Response Is Associated with an
Increase in ALA Synthesis
The FR block of greening has previously been ascribed, at least in
part, to the photoexcitation of excess Pchlide in the absence of
adequate POR buffering (Runge et al., 1996; Sperling et al., 1997), and
we hypothesized that the WL-dependent, transcriptionally uncoupled
response was likely to be related to these processes. ALA synthesis is
the rate-limiting step in Pchlide accumulation (Beale and Castelfranco,
1974), and GluTR activity and HEMA1 expression are believed
to be key determinants of this activity (see McCormac et al., 2001).
Because FRc resulted in an increase in HEMA1 expression (Fig. 3, A and B), we tested whether HEMA1 induction could
also contribute to the transcriptionally uncoupled response by
increasing ALA synthesis and, therefore, the pool of photosensitive Pchlide.
Figure 4 shows that irradiation with FRc significantly elevated ALA
synthesis in WT seedlings (Fig. 4A) and this was correlated (r = 0.82) with HEMA1 mRNA levels under a
range of different conditions (Fig. 4B). On 3% (w/v) Suc, the FRc
induction of ALA synthesis was also seen, consistent with the
observation previously (Fig. 2) that although Suc inhibits the
transcriptionally coupled response, the transcriptionally uncoupled
response is retained. In phyA seedlings, the FR-mediated
increase in ALA synthesis was significantly reduced on 0% and 3%
(w/v) Suc (Fig. 4A). The greater increase in ALA synthesis in WT
seedlings grown under FRc also correlated with a higher level of
Pchlide in WT versus phyA seedlings in FRc (Fig. 4A, inset).
In contrast to the WT response, phyA seedlings showed no
correlation between ALA synthesis rates and HEMA1 mRNA levels (Fig. 4B, inset), indicating that the FR-mediated increase in
ALA synthesis in phyA was not attributable to
transcriptional regulation of HEMA1. This strongly suggests
a role for posttranscriptional effects on GluTR expression and/or
activity in regulating ALA synthesis under these conditions.
We next investigated the relationship between ALA
synthesis rates and subsequent greening capacity. Figure 4C shows this
data for WT and phyA lines grown under conditions in which
only the transcriptionally uncoupled response is occurring (i.e. 1-3 d of FRc with Suc or 1 d of FRc without Suc). There is a
significant (r = 0.88) inverse relationship between ALA
synthesis rates and the ability to green following transfer to WL. This
suggests that increased ALA synthesis contributes to the
transcriptionally uncoupled response.
Loss of Nuclear Gene Expression following FRc- and Norflurazon
(NF)-Induced Photobleaching Is Additive
The loss of nuclear gene expression in the transcriptionally
coupled FR block of greening response is reminiscent of the loss of
HEMA1 expression following plastid photobleaching (McCormac et al., 2001). Therefore, we investigated the interaction between these
two responses. Figure 5A shows that the
carotenoid biosynthesis inhibitor NF inhibits HEMA1
expression under WL following growth in darkness on 0% or 3% (w/v)
Suc. This response is the same in WT and phyA seedlings
(Fig. 5, A and B). NF does not affect the HEMA1 response of
etiolated seedlings while under FR (data not shown). However, when NF
and a 3-d FR pretreatment were combined on 0% (w/v) Suc, the
inhibitory effect on the subsequent WL responsiveness of the
HEMA1 promoter exceeded that of either of the individual treatments (Fig. 5A). It should be noted that immediately prior to
transfer to WL, the FR-irradiated seedlings in each case had higher
levels of HEMA1 expression than dark controls and so had the
potential to show a higher level of transcriptional inhibition under
WL. However, direct observation of HEMA1 mRNA levels by northern blotting demonstrates that in the presence of NF, a
substantial level of HEMA1 mRNA remains (data not shown).
Therefore, the magnitude of the NF-mediated reduction of GUS activity
was not limited following darkness, and only the combination of FR and
NF produced the minimum promoter activity (Fig. 5A). The additive
transcriptional effect of NF and an FR pretreatment was dependent on
phytochrome A (Fig. 5B). In contrast, the transcriptionally uncoupled
response, as seen in WT seedlings grown on 3% (w/v) Suc (Fig. 2), did
not enhance the NF-mediated inhibition of the HEMA1 WL
response (Fig. 5A). Therefore, these results show that the loss of
nuclear gene expression during the FRc-mediated transcriptionally
coupled response and following NF-induced photobleaching is additive.
This may result from two inputs into the same plastid signaling pathway
or through two independent pathways.
View larger version (35K):
[in this window]
[in a new window]
|
Figure 5.
POR overexpression rescues the loss of
HEMA1 expression in WL following growth in FRc or NF. A,
Effect of a combined FR preirradiation and NF treatment on the WL
response of a HEMA1 promoter:gusA reporter gene.
WT seedlings were grown with or without 3% (w/v) Suc for 3 d in
FRc or darkness prior to transfer to WL (130 µmol
m2 s1) for an
additional 3 d. NF was included in the media at 5 µM and resulted in the cotyledons appearing
completely white under all treatments. Values were normalized to the
respective dark control levels (=100) within each experiment. Data
shown are the mean ± SE (n = 5). B, HEMA1 promoter:gusA expression in
phyA seedlings under the same conditions as shown in A. C,
The effect of POR overexpression on HEMA1 expression.
Seedlings of PORA- or PORB-overexpressors (PAO-3 and PBO-1) were
germinated in the dark without Suc and were irradiated for 0 to 3 d under FRc. HEMA1 mRNA accumulation was measured following
transfer to WL (130 µmol m2
s1). D, The effect of 5 µM NF on the accumulation of HEMA1
mRNA in WL-irradiated seedlings of WT and transgenic lines
overexpressing PORA (PAO-3) or PORB (PBO-1). Seedlings were grown in
the absence of Suc for 3 d in the dark prior to transfer to WL for
an additional 3 d.
|
|
POR Overexpression Leads to a Maintenance of Nuclear Gene
Expression following FRc- and NF-Induced Photobleaching
It has previously been shown that overexpression of POR can
partially rescue the FR block of greening response, and this has been
attributed to an inhibition of WL intensity-dependent photooxidative damage (Runge et al., 1996; Sperling et al., 1997). We wanted to
examine whether POR overexpression could also affect the
transcriptionally coupled response and, therefore, we subjected
transgenic Arabidopsis lines overexpressing PORA (PAO-3) or PORB
(PBO-1; Sperling et al., 1997) to conditions resulting in the loss of
HEMA1 in WT seedlings (i.e. 2-3 d of FRc on 0% [w/v]
Suc). Upon transfer to WL, FR-treated PAO-3 and PBO-1 seedlings
retained full expression of HEMA1 (Fig. 5C). Thus, POR
overexpression inhibited the transcriptionally coupled FR block of
greening response. The plastid localization of POR in the
overexpressing lines (Sperling et al., 1997, 1999) supports the idea
that loss of HEMA1 expression following FRc is the
consequence of a signal emanating from the plastid.
We also tested the effect of POR overexpression on the loss of plastid
signaling following NF treatment. The cotyledons of NF-treated
seedlings of WT and both POR-overexpressing lines appeared totally
white. However, as seen for the WL transcriptional response following
FRc, the POR-overexpressing seedlings accumulated significant levels of
HEMA1 mRNA (Fig. 5D), indicating that positive plastid signaling was still functional despite photobleaching of the
cotyledons. Therefore, POR overexpression results in a phenotype that
is apparently similar to that of the genomes uncoupled
mutants in which nuclear gene expression is maintained in the absence
of functional plastids (Susek et al., 1993). However, in this case, it
is not yet known whether POR overexpression results in a reduced level
of plastid damage or has a direct affect on plastid signaling.
The Transcriptionally Uncoupled But Not the Transcriptionally
Coupled FR Block of Greening Response Is Inhibited by Cytokinin
We have described the FR block of greening as comprising two
separate responses with respect to their different sensitivities to WL
intensity and Suc and their effects on nuclear gene expression. Furthermore, there is a time-dependent transcription effect that is superimposed on these processes. Thus, the FR block of greening is a
complex process encompassing a phytochrome A-dependent FR-HIR and
phytochrome-independent effects. To try to isolate these
components further, we investigated the effects of cytokinin treatment,
as this hormone has previously been shown to influence greening
capacity (e.g. Kusnetsov et al., 1998).
WT and phyA seedlings grown for 5 d in the dark in the
absence of Suc showed a time-dependent loss of light responsiveness (see Fig. 3) and, therefore, showed a characteristically poor transcriptional response of the HEMA1 promoter to a WL
treatment (Fig. 6A) and a partially
reduced greening response (Fig. 6C). Application of the cytokinin
6-benzylaminopurine (BAP) inhibited both of these time-dependent
processes (Fig. 6, A and C). In contrast, the inhibition of
promoter responsiveness and cotyledon greening seen in WT seedlings
subject to FRc (the transcriptionally coupled response) was not
prevented by cytokinin (Fig. 6, A and C). Analysis of HEMA1
expression in darkness or FR light showed that cytokinin increased
HEMA1 expression by an equal increment in both conditions (Fig. 6E) and, therefore, the FR response relative to darkness was
unchanged.
View larger version (42K):
[in this window]
[in a new window]
|
Figure 6.
The effect of cytokinin on the FR block greening
response and HEMA1 expression. A and B, The effect of
cytokinin on WL-mediated induction of HEMA1 expression. WT
and phyA seedlings were germinated in the dark for 2 d,
transferred to FRc (or maintained in the dark) for 3 d, and then
transferred to WL for an additional 3 d prior to measurement of
GUS activity. Seedlings were grown in the absence (A) or presence (B)
of 3% (w/v) Suc, and data shown are the mean ± SE (n = 3). C and D, The effect
of cytokinin on greening following transfer to WL. Seedlings were grown
as shown in A and B and greening was measured in the absence (C) or
presence (D) of 3% (w/v) Suc. Data shown are the mean ± SE (n = 3). E and F, The effect
of cytokinin on HEMA1 expression in the dark and FRc. WT and
phyA seedlings were germinated in the dark for 2 d and
were transferred to FRc (or maintained in the dark) for 3 d prior
to measurement of GUS activity. Seedlings were grown in the absence (E)
or presence (F) of 3% (w/v) Suc, and data shown are the mean ± SE (n = 3).
|
|
When seedlings were grown on 3% (w/v) Suc, an FR pretreatment had
little affect on the subsequent HEMA1 response to WL (Fig. 6B), as seen previously (Fig. 2), and cytokinin did not further influence this (Fig. 6B). For seedlings grown in darkness on 3% (w/v)
Suc, the greening capacity was very slightly reduced by the presence of
cytokinin (Fig. 6D). However, cytokinin could almost completely prevent
the transcriptionally uncoupled block of greening response following FR
irradiation of WT seedlings (Fig. 6D). Again, cytokinin stimulated
HEMA1 expression in dark-grown seedlings (Fig. 6F), but in
contrast to the effect on 0% (w/v) Suc (Fig. 6E), cytokinin inhibited
the FR-mediated increase in HEMA1 expression on 3% (w/v)
Suc (Fig. 6F).
These results demonstrate that exogenous cytokinin can rescue the
transcriptionally uncoupled response as it develops in the presence of
Suc, but that it does not block the transcriptionally coupled response.
However, there is also an effect of cytokinin on the time-dependent,
FR-independent inhibition of greening. Therefore, the response to
cytokinin defines a developmental separation of the two FR block of
greening responses and can also isolate the FR-mediated
transcriptionally coupled response from FR-independent events.
| |
DISCUSSION |
Two Distinct Responses Leading to an FR Block of
Greening
Here, we demonstrate that the FR block of greening comprises two
distinct FR-dependent responses. As shown in Figure
7, the two responses can be distinguished
by a number of different parameters. Following the onset of FRc
irradiation, the first response detected (within 1 d of FRc) is a
WL intensity-dependent incomplete loss of greening ability with a
retention of WL induction of HEMA1 and Lhcb
expression. We have shown that this response, which we have termed the
transcriptionally uncoupled response, is prevented by the
phyA mutation and by cytokinin treatment. In addition, it
has previously been demonstrated that overexpression of POR can inhibit
this response (Sperling et al., 1997). Following longer periods of FRc
irradiation (3 d of FRc), a WL intensity-independent response is also
observed (Fig. 7). This response is characterized by a complete loss of
greening ability and the inhibition of HEMA1 and
Lhcb expression following transfer to WL, and is inhibited by Suc and POR overexpression. We have termed this response the transcriptionally coupled response and it is also absent in the phyA mutant, consistent with the FR fluence rate dependence
of this process. We have also identified a time-dependent effect on
greening and the transcriptional light response that proceeds in the
absence of FR.
View larger version (18K):
[in this window]
[in a new window]
|
Figure 7.
Model for the progression of the FR block of
greening response. FR block of greening is comprised of two separate
phytochrome A-dependent responses that can be separated based on their
dependence on WL intensity and their effects on greening and nuclear
gene expression. The transcriptionally uncoupled response is WL
intensity dependent, results in a partial loss of greening following
transfer to WL, and is inhibited by cytokinin, the phyA
mutation, and POR overexpression. The transcriptionally coupled
response is WL intensity independent, and results in a complete loss of
greening ability and an inhibition of HEMA1 and
Lhcb gene expression in WL. This response is inhibited by
Suc, the phyA mutation, and POR overexpression and requires
high fluence rate FRc. 1Sperling et al.
(1997).
|
|
The Transcriptionally Uncoupled Response Results from a
Perturbation of the PORA:Pchlide Ratio Leading to Increased
Photosensitivity to WL
Previous studies on the FR block of greening response have
demonstrated that there is an FR-HIR-mediated depletion of PORA in WT
seedlings (Barnes et al., 1996; Runge et al., 1996) and that this leads
to an overaccumulation of nonphotoactive Pchlide (Lebedev et al.,
1995; Runge et al., 1996). This Pchlide species cannot be reduced to
Chlide upon illumination and therefore acts as a potent sensitizer for
photooxidative damage to plastids (Reinbothe et al., 1996). The
combination of photooxidative damage and reduced availability of POR
for chlorophyll synthesis leads to a loss of greening ability in WL.
Consistent with this interpretation, overexpression of POR has been
shown to protect against the FR block of greening by decreasing the
amount of nonphotoactive Pchlide present in seedlings prior to the
shift to WL (Sperling et al., 1997, 1999). Because the effects of
photooxidative damage are dependent on WL intensity, we believe that
the transcriptionally uncoupled response characterized in this study
corresponds to an increase in photosensitivity through inhibition of
POR synthesis as shown previously (Barnes et al., 1996; Runge et al.,
1996; Sperling et al., 1997). One aspect of the block of greening
response is that there is an FR-dependent increase in total
Pchlide (Fig. 4; Sperling et al., 1997, 1999). This would be expected
to compound any depletion in POR. Here, we show that the increase in
Pchlide results from an FR-HIR-mediated up-regulation of
HEMA1 expression and ALA synthesis (Fig. 4). The correlation
between increased ALA synthesis and loss of greening ability under
conditions resulting in the transcriptionally uncoupled FR block of
greening response (Fig. 4) supports the hypothesis that the
transcriptionally uncoupled response is a consequence of deregulation
of the tetrapyrrole pathway (Fig. 7).
The transcriptionally uncoupled response was inhibited by the presence
of cytokinin. The effect of cytokinin in inhibiting the greening
response of FR-treated WT seedlings grown on 3% (w/v) Suc (Fig. 6D)
corresponded to a decrease in the FR-mediated induction of
HEMA1 expression (Fig. 6F). Cytokinin has also been shown to increase PORA levels (Kusnetsov et al., 1998), and these combined effects could well be sufficient to maintain adequate buffering of
Pchlide when exposed to WL. It should also be noted that in addition to
blocking the transcriptionally coupled response, Suc slightly delayed
the transcriptionally uncoupled response (Fig. 2A). Again, this
corresponded to a delay in the induction of HEMA1 (Fig. 3D)
and in the down-regulation of POR expression (Barnes et al.,
1996). These results are all consistent with the PORA:Pchlide ratio
being the determining factor in the transcriptionally uncoupled FR
block of greening response.
With regard to the physiological significance of these results, the
observation here that a 1-d FR pretreatment (or longer irradiations
with low-intensity FRc) could enhance the subsequent transcriptional
response to low-intensity WL (Fig. 1C) has some precedence in reports
that an FRc pretreatment caused an amplification of the low fluence
growth response to a subsequent RL treatment (Casal, 1995). Stimulation
of ALA synthesis in WL through an FR pretreatment has also been noted
before (e.g. Masoner and Kasemir, 1975). Therefore, it appears that an
FR-HIR serves a similar function as a R light pulse (followed by a
short darkness incubation period) in priming the autotrophic transition
in the emerging seedling, in this case, for example, under dense canopy shade.
The Transcriptionally Coupled Response Results in the Loss of
Nuclear Gene Expression through an FR-HIR Effect on Plastid Gating of
Phytochrome Signaling
The transcriptionally coupled response was characterized by
the almost complete loss of HEMA1 and Lhcb
expression in WL following transfer from prolonged (>2 d) FR
treatments (Fig. 1). Because this effect is also seen using the
HEMA1 promoter:gusA lines, it clearly results
from a loss of promoter-driven transcriptional activity as opposed to
transcript stability. Suc (Fig. 2), but not cytokinin (Fig. 6),
inhibited the development of the transcriptionally coupled response. We
propose that the loss of nuclear gene expression results from an
FR-HIR-mediated inhibition of a plastid signal that gates
HEMA1 and Lhcb gene expression. This proposal is
based on the similarity of this response to the characteristic loss of
nuclear gene expression following NF-induced photobleaching and the
ability of POR overexpression to rescue HEMA1 expression (Fig. 5). Because POR is localized to plastids in these overexpressors (Sperling et al., 1997, 1999), it is reasonable that its ability to
rescue HEMA1 expression results from alterations in plastid function.
It is not clear whether the deterioration of the plastid signal in the
transcriptionally coupled response occurs directly during the FR
irradiation period or subsequently under WL as the result of
sensitization by the FR pretreatment. However, our current hypothesis
is that plastid signaling is lost during the period of FRc irradiation.
This hypothesis is based on the observations that the complete loss of
greening ability and nuclear gene expression was independent of WL
intensity (Fig. 1), but was dependent on the FR fluence rate (Fig. 3F).
Furthermore, there was an additive interaction of an FR pretreatment
with NF-induced photobleaching (Fig. 5), which suggests that plastid
signaling is not completely abolished by either treatment alone.
Although it is not possible to say whether these two treatments use the
same or independent signaling pathways, these results do suggest that
at least some element of the transcriptional block was irreversibly
determined during the FR preirradiation period. This is in contrast to
the effect of NF-induced photobleaching, which occurs exclusively during the WL irradiation period.
A number of events may be occurring during the period of FRc that could
lead to the loss of plastid signaling. Previous studies on the effects
of FR irradiation on plastid structure have revealed that this
treatment leads to a deficiency in the ordered membrane system of the
prolamellar body (Oelmüller et al., 1986; Barnes et al., 1996;
Sperling et al., 1997). This can be largely attributed to the dramatic
decline in PORA (Runge et al., 1996), which is the major protein
component of the prolamellar body (Ryberg and Sundqvist, 1982; Ikeuchi
and Murakami, 1983). In addition, the development of vesicles, which
may be related to degradation of plastid components, has also been
observed (Barnes et al., 1996). These ultrastructural changes all occur
during the FR treatment, prior to transfer to WL, and seedlings treated
with 2 d of darkness following the FR pretreatment do not recover
greening ability (Barnes et al., 1996), indicating that FRc leads to an
irreversible arrest of plastid development. It is interesting that Suc,
which inhibits the WL-independent loss of nuclear gene expression, also inhibits vesicle formation (Barnes et al., 1996). It is plausible that
this FR-mediated deterioration of plastid structure would compromise
plastid/nuclear signaling and is consistent with the observation here
of a specific FR-HIR-mediated impairment of nuclear gene expression.
Because POR overexpression rescues the observable effects of FRc on
plastid ultrastructure (Sperling et al., 1997, 1999), the demonstration
that POR overexpression also rescues HEMA1 expression
supports the idea that FRc-mediated changes in plastid ultrastructure
lead to a loss of plastid signaling. In this context, it may be that
cytokinin can mediate plastid repair (sufficient to permit greening) in
the event of low-level damage such as within the transcriptionally
uncoupled response, but cannot overcome this more severe damage
associated with the transcriptionally coupled response.
Loss of nuclear gene expression can also be seen following transfer to
WL after longer growth periods (>4 d) in darkness (Fig. 3). In this
experiment, Arabidopsis seedlings showed a progressive loss of the
HEMA1 promoter response to light and this was independent of
FR and phytochrome. Lhcb and RBCS mRNA levels
have previously been reported to be subject to light-independent
developmental mechanisms (Brusslan and Tobin, 1992), but it is equally
possible that this is simply the result of starvation. The
inhibition of this response by 3% (w/v) Suc supports such an idea. The
decline in the transcriptional response of seedlings while still under FRc displayed a similar time course to the FR-independent loss of WL
induction, and this response was also prevented by the presence of 3%
(w/v) Suc (Fig. 3). However, the time-dependent loss of transcriptional
activity is distinct from that seen following the complete FR block of
greening response. First, the use of the GUS reporter showed that an FR
pretreatment resulted in a significantly greater level of
HEMA1 promoter inhibition under subsequent WL than could be
accounted for by an extended dark period (Fig. 1C). Second, the
time-dependent loss of greening ability and promoter responsiveness, as
seen in dark-grown seedlings in the absence of Suc, was alleviated by
cytokinin, whereas the FR-dependent loss of greening and
HEMA1 expression was not (Fig. 6, A and C). The most likely
explanation is that in the absence of Suc or cytokinin, these time- or
starvation-dependent effects proceed within FR-irradiated seedlings in
parallel with FR-specific responses. In contrast, when seedlings are
transferred to R or WL, they become dissociated from such a
time-dependent loss of transcriptional activity (Fig. 3C) through the
photosynthetic supply of sugars or direct chloroplast signaling.
There is clearly a close link between the transcriptionally uncoupled
response proposed to act through a perturbation of the PORA:Pchlide
ratio and the transcriptionally coupled response in which it is
proposed that irreversible changes in plastid ultrastructure result in
a loss of nuclear gene expression. The important role of POR in
determining plastid structure and the ability of POR overexpression to
rescue both responses, at least in part (Fig. 5; Sperling et al.,
1997), supports such a hypothesis and suggests that the effects caused
by the transcriptionally uncoupled response may have an input into the
transcriptionally coupled response (Fig. 7). Given the temporal
separation of the two FR-mediated responses, it is possible to think of
them proceeding sequentially following the onset of FRc, with the
transcriptionally uncoupled response leading to the transcriptionally
coupled response. However, as cytokinin inhibits the transcriptionally
uncoupled response while having no effect on the transcriptionally
coupled response, there is clearly no absolute requirement for the
transcriptionally uncoupled response for the transcriptionally coupled
response to proceed.
Implications for Plastid-Nuclear Signaling
In this study, we have identified a plastid-signaling pathway that
is affected by prolonged periods of FRc irradiation and is at least
partially independent from the pathway inhibited by NF-induced
photobleaching. Previous studies have also concluded that more than one
plastid-signaling pathway must exist (Vinti et al., 2000;
Mochizuki et al., 2001). There are a number of similarities between the
FRc dependence of plastid signaling described here and previous
observations. For example, treatment of barley (Hordeum vulgare) seedlings with the carotenoid inhibitor Amitrole
(but not NF) results in the loss of Lhcb and RbcS
expression in low-intensity, nonphotodamaging WL (La Rocca et al.,
2001). In a similar manner, inhibition of etioplast development
resulting from the reduced accumulation of Pchlide and POR in the
phytochrome chromophore-deficient aurea mutant of tomato
(Terry and Kendrick, 1999; Terry et al., 2001) also leads to reduced
Lhcb expression in the darkness (Sharrock et al., 1988;
Ken-Dror and Horwitz, 1990). This can also be considered as a WL
intensity-independent loss of plastid signaling. The possibility that
changes in plastid development through FRc, inhibition of chromophore
biosynthesis, and Amitrole treatment all lead to inhibition of the same
plastid signaling pathway is intriguing and will require further
experiments. One common feature of these conditions is that they all
perturb tetrapyrrole biosynthesis (Terry and Kendrick, 1999; La Rocca
et al., 2001). This observation, together with our finding that POR
overexpression can sustain this signaling response, supports the
growing evidence that tetrapyrroles play a major role in mediating
plastid signaling (Kropat et al., 2000; Vinti et al., 2000;
Mochizuki et al., 2001; Møller et al., 2001). However, the exact role
of tetrapyrroles in these pathways is still unknown and further work is
clearly needed to elucidate this. The identification of a new pathway
leading to a phytochrome A-dependent loss of plastid signaling will
provide a new experimental system for addressing questions on the
mechanisms involved in plastid-nuclear signaling.
| |
MATERIALS AND METHODS |
Plant Material
Transgenic lines expressing a
HEMA1promoter:gusA reporter construct
were the kind gift of Andrea Fischer and Prof. Dieter Söll (Yale
University, New Haven, CT) and have been described previously (McCormac
et al., 2001). Two homozygous lines were used in these experiments,
each containing the full-length (
2,435/+252) HEMA1
promoter region fused upstream of the GUS coding sequence. The
Arabidopsis WT ecotypes Col and Ler and the
phyA (Col and Ler) mutant used in this
study were kindly provided by Drs. Haruko Okamoto and Xing-Wang Deng
(Yale University).
Surface-sterilized seeds were plated onto a 1% (w/v) agar medium
containing Murashige and Skoog salts (Murashige and Skoog, 1962) and
were supplemented with 0% or 3% (w/v) Suc as indicated in
"Results." Plates were placed at 4°C in darkness for 2 d
prior to receiving a 30-min WL irradiation to synchronize germination. Unless indicated otherwise, seeds were allowed to germinate in darkness
at 23°C for 2 d (0% [w/v] Suc) or 3 d (3% [w/v] Suc) prior to transfer to the FR light source. Where indicated, NF (kindly
provided by Prof. John Gray, University of Cambridge, Cambridge, UK)
was added to the medium at a level of 5 µM. For treatments with the cytokinin BAP, seeds were first germinated on
sterile filters placed over Murashige and Skoog medium for 1 d in
darkness and were then transferred to plates containing 10 mg
L1 (44 µM) BAP.
Light Sources
Broad-band WL was provided by white fluorescent tubes (400-700
nm = 130 µmol m2 s1 unless indicated
otherwise). This fluence rate was equivalent to high-intensity WL (250 µmol m2 s1; Figs. 1 and 2) for the
greening response. Narrow waveband sources were provided by
light-emitting diode displays in environmental control chambers
(Percival Scientific, Boone, IA). R light corresponded to a peak at 669 nm (25-nm bandwidth at 50% of peak magnitude) with a fluence rate of
80 µmol m2 s1. FR from the light-emitting
diodes had a peak at 739 nm (25-nm bandwidth at 50% of peak magnitude)
and was passed through two filters (nos. 116 and 172; Lee Filters,
Andover, UK) to remove 
s <700 nm to give a final fluence rate of 10 µmol m2 s1.
RNA Gel-Blot Analysis
RNA gel-blot analysis was performed using total RNA (30 µg
lane1) exactly as described previously (McCormac et al.,
2001). The HEMA1 probe used was a 3'-cDNA fragment that
is gene specific (Kumar et al., 1996; McCormac et al., 2001). The
Lhcb probe (kindly provided by Dr. Joanne Chory, The
Salk Institute, La Jolla, CA) contained the majority of the
coding region of the Lhcb1*2 gene and is predicted to
crosshybridize with other members of the Lhcb gene
family (McCormac et al., 2001).
GUS Fluorometric Analysis
Quantitative assays of GUS activity in seedling cotyledons were
conducted exactly as described previously (McCormac et al., 2001).
ALA Biosynthesis Assay
Seedlings were incubated, under the appropriate light
conditions, in 100 mM sodium phosphate buffer (pH 7.0),
with or without 40 mM levulinic acid (Sigma-Aldrich, Poole,
UK), at 23°C for 7 h. Seedlings were then blotted dry, frozen in
liquid nitrogen, and stored at 80°C. The frozen seedlings were
ground in 20 mM sodium phosphate buffer (pH 7.0), incubated
on ice for 20 min, and centrifuged at 8,500g in a
benchtop microfuge for 5 min. The supernatant was incubated with
ethylacetoacetate at 100°C followed by the addition of modified
Ehrlich's reagent (Urata and Granick, 1963). Absorbance was read at
526, 553, and 600 nm and the concentration of ALA was calculated using
a molar absorption coefficient of 7.45 × 104
M1 cm1.
Pigment Extraction
Pchlide was extracted based on the method of Rebeiz et al.
(1975). Etiolated Arabidopsis seedlings were homogenized in acetone:0.1 M NH4OH (90:10, v/v). The extract was
centrifuged at 8,500g for 2 min and the pellet was
re-extracted as above. The supernatants were combined and mixed with an
equal volume of hexane. The aqueous and hexane fractions were collected
separately, and relative fluorescence emission spectra were recorded
using a fluorescence spectrophotometer (F-2000; Hitachi, Tokyo) with an
excitation wavelength of 440 nm. Greening is shown as the percentage of
seedlings with visibly green cotyledons. Comparison with direct
measurements of chlorophyll extractions (Moran, 1982) showed that such
an assessment was linearly correlated to total chlorophyll within a
given genetic background.
We thank Andrea Fischer and Prof. Dieter Söll (Yale
University) for the HEMA1 promoter:gusA
lines, Dr. Gregory Armstrong (Ohio State University, Columbus) for the
PORA and PORB overexpressing lines, Dr. Xing-Wang Deng (Yale
University) for seeds of phyA and hy1,
Prof. John Gray (University of Cambridge) for the generous gift of
Norflurazon, and Drs. Haruko Okamoto (University of Oxford), James
Weller (University of Tasmania, Hobart, Australia), and Enrique
López-Juez (Royal Holloway College, University of London) for
critical reading of the manuscript.
Received February 5, 2002; returned for revision April 14, 2002; accepted May 5, 2002.
Article, publication date, and citation information can be found at
www.plantphysiol.org/cgi/doi/10.1104/pp.003806.
TOP
ABSTRACT
INTRODUCTION
-aminolevulinic acid in higher plants: accumulation of 
-aminoketones and the metabolism of aminoacetone.
J Biol Chem
238: 811-820
